Effect of the Insertion of a Glycine Residue into the Loop Spanning

Dec 4, 2008 - To whom correspondence should be addressed. Telephone: (614) 292-9428. Fax: (614) 292-6773. E-mail: [email protected]., ‡. Ohio State ...
0 downloads 0 Views 2MB Size
Download PDF
13788

Biochemistry 2008, 47, 13788–13799

Effect of the Insertion of a Glycine Residue into the Loop Spanning Residues 536-541 on the Semiquinone State and Redox Properties of the Flavin Mononucleotide-Binding Domain of Flavocytochrome P450BM-3 from Bacillus megaterium Huai-Chun Chen‡ and Richard P. Swenson*,‡,§ Department of Biochemistry and Ohio State Biochemistry Program, The Ohio State UniVersity, Columbus, Ohio 43210 ReceiVed May 20, 2008; ReVised Manuscript ReceiVed NoVember 7, 2008

ABSTRACT:

Despite sharing sequence and structural similarities with other diflavin reductases such as NADPH-cytochrome P450 reductase (CPR) and nitric oxide synthase, flavocytochrome P450BM-3 displays some unique redox and electron transferring properties, including the inability to thermodynamically stabilize the neutral semiquinone (SQ) state of the flavin mononucleotide (FMN) cofactor. Rather, the anionic SQ species is only transiently formed during rapid reduction. Why is this? The absence of a conserved glycine residue and, as a consequence, the shorter and less flexible cofactor-binding loop in P450BM-3 represents a notable difference from other diflavin reductases and the structurally related flavodoxin. This difference may facilitate the formation of a strong hydrogen bond between backbone amide NH group of Asn537 and N5 of the oxidized FMN, an interaction not found in the other proteins. In the flavodoxin, the conserved glycine residue plays a crucial role in a redox-linked conformational change that contributes to the thermodynamic stabilization of the neutral SQ species of the FMN through the formation of a hydrogen bond with the N5H group of the flavin. In this study, a glycine residue was inserted after Tyr536 in the loop within the isolated FMN-binding domain as well as the diflavin reductase domain of P450BM-3, a position equivalent to Gly141 in human CPR. As a result, the insertion variant was observed to accumulate the neutral form of the FMN SQ species much like CPR. The midpoint potential for the SQ/HQ couple decreased by 68 mV, while that for the OX/SQ couple remained unchanged. 15N NMR data provide evidence of the disruption of the hydrogen bond between the backbone amide group of Asn537 and the N5 atom in the oxidized state of the FMN. Molecular models suggest that the neutral FMN SQ could be stabilized through hydrogen bonding with the backbone carbonyl group of the inserted glycine residue in a manner similar to that of CPR and the flavodoxin. The insertion of the glycine at the same location within the diflavin domain resulted in a purified protein that retained nearly stoichiometric levels of bound FAD but tended to lose the FMN cofactor. This preparation retained one-third of the ferricyanide reductase activity but 65 mV to a more negative value of -245 mV compared that of the wild type (-177 mV) (Table 1). Interestingly, the midpoint potential for the OX/SQ couple in the G537ins variant remained significantly lower than for the human CPR (-43 mV), while ESQ/HQ became more comparable (-280 mV) (11). The potential difference between the one-electron couples of ∼47 mV in the G537ins variant was in close agreement with that calculated from the stability constant (Ks) for the neutral SQ based on its accumulation (∼55%) in both in the reductive and oxidative titrations and the following relationship: Ks )

[FBDSQ]2 F o o - E SQ ) exp (E ⁄ HQ) [FBDOX][FBDHQ] RT OX ⁄ SQ

[

] (2)

The extent of accumulation of the neutral SQ and the difference between the midpoint potentials of the two couples

are substantially smaller than for those observed in human CPR and C. beijerinckii flavodoxin, ∼240 and ∼ 300 mV, respectively (11, 17). Binding of the FMN to G537ins in Each Oxidation State. A feature of wild-type FBDBM3 is the rather modest difference in the binding of each of the three oxidation states of the cofactor compared to those of many of the other FMNbinding domains. Binding free energies of -10.1, -10.8, and -10.7 kcal/mol were established for the OX, SQ, and HQ states, respectively (Table 1). The dissociation constant (Kd) for oxidized FMN for the G537ins variant was determined directly by spectrophotometric titrations to be 408 nM. This represents an ∼10-fold increase compared to that for wild-type FBDBM3 determined here under similar conditions (41 nM) and the published value of 31 nM from Haines et al. (26). This change represents a loss of ∼1.4 kcal/mol of binding free energy. The Kd values for the SQ and HQ states, which cannot be established directly, were calculated from the experimental values for the midpoint potentials for each couple and the Kd for the oxidized FMN based on a linked equilibrium analysis (27). The values were determined to be ∼7- and ∼105-fold higher for the SQ and HQ states, respectively, than for wild-type FBDBM3 (Table 1). The free energy of binding of the FMN in the SQ state was increased by 0.9 kcal/mol relative to the OX state for the insertion variant. The FMNHQ complex was substantially less stable (by ∼2.7 kcal/mol) than the wild-type protein and ∼1.6 kcal/ mol less stable than the SQ complex in G537ins. This observation contrasts with that for the wild type, for which there is little difference between the binding of the SQ and HQ species. 15 N NMR Studies of G537ins. Previous studies of the flavodoxin have shown that while the N5 atom of the FMN is not hydrogen bonded in the oxidized state, a redox-linked conformation change permits the formation of such a bond between a protein backbone carbonyl group and the N5H group of the flavin in both the neutral SQ and HQ states (17). This interaction results in the thermodynamic stabilization of neutral FMNSQ and the higher potentials for the OX/ SQ couple that characterize these proteins. X-ray crystallographic and NMR spectroscopic studies support a similar peptide backbone-mediated stabilization of these redox states in human CPR (8, 19). In contrast, a strong hydrogen bond between the N5 atom of the flavin and the amide NH group of the peptide backbone is already established in the OX state in wild-type FBDBM3 (9). With those observations in mind, 1D 15N NMR studies were initiated to evaluate the effect of the glycine insertion on the changes in hydrogen bonding and the flavin environment in both OX and HQ states of the FBDBM3. The 1D 15N NMR spectrum of apoG537ins reconstituted with [15N]FMN in the OX state is shown in Figure 4 along with the wild-type FBDBM3 spectrum for comparison. Three resonance peaks were observed for the G537ins variant (Figure 4A). The 15N chemical shifts were assigned for the N1, N3, and N10 atoms by comparison to wild-type FBDBM3 (21) (Figure 4B and Table 2). The resonance for the N5 atom was not as clearly established. A weak signal was occasionally observed in the 310 ppm region; however, the signal-to-noise ratio in this region was poor, and this chemical shift value was outside the typical range for the N5 atom in aqueous and apolar environments (35). It is also possible that the N5 resonance has been

Glycine Insertion Modulates the Redox Properties of P450BM-3

Biochemistry, Vol. 47, No. 52, 2008 13793

Table 1: One-Electron Midpoint Potentials, Dissociation Constants, and Binding Energies for Each Oxidation States of the FMN Cofactor in Wild-Type FBDBM3 and G537ins Kd (µM)b

midpoint potential (mV) EOX/SQ

ESQ/HQ

OX

SQ

HQ

-206 -198 ( 11

-177 -245 ( 14

0.041 ( 0.018 0.41 ( 0.09

0.012 0.086

0.014 1.47

a

wild type G537ins

Gibbs free energy (kcal/mol)

a

OX

∆G

∆GSQ

∆GHQ

-10.1 -8.7

-10.8 -9.6

-10.7 -8.0

a From ref 22. b Kd values in the OX state were determined by spectrophotometric titration of FMN with apoprotein, and those in the SQ and HQ states were calculated as described in Experimental Procedures (27).

FIGURE 4: (A) 1D 15N NMR spectra of the G537ins variant reconstituted with 15N-enriched FMN in the oxidized state. The 15N chemical shift values of 187.1, 159.8, and 157.9 ppm for the three resonance peaks were assigned to the N1, N10, and N3 atoms of the flavin, respectively. A signal was occasionally observed at ∼310 ppm as shown in this spectrum but was not conclusively assigned (see Results). For comparison, the 15N chemical shift values assigned for the N5, N1, N10, and N3 atoms for the FMN bound to wild-type FBDBM3 are 321.5, 189.0, 162.6, and 160.5 ppm, respectively (B) (21). Both spectra were recorded under similar conditions with the proteins dissolved in 100 mM sodium phosphate buffer (pH 7.0) containing 10% D2O, and using [15N]urea as the external standard reference. Table 2: 15N Chemical Shift Values for Free and Bound FMN in the Oxidized State 15

N NMR chemical shift (ppm)

atom

FMN

TARFa

wild-type FBDBM3b

G537ins

N1 N3 N5 N10

190.8 160.5 334.7 164.6

199.9 159.8 344.3 150.2

189.0 160.5 321.5 162.6

187.1 157.9 (309.8) 159.8

a

a

From ref 35. TARF, tetraacetylriboflavin in CHCl3. b From ref 21.

broadened beyond detection due to conformational flexibility introduced into this region by the insertion, with the resonance at ∼310 ppm representing an artifact. Unfortunately, the extension of data recording times beyond 36 h to improve the signal-to-noise ratio in this region resulted in protein precipitation. Nonetheless, the data suggest that the hydrogen bond and the environment of the N5 atom of the FMN have been significantly perturbed in the G537ins variant. As for wild-type FBDBM3, the 15N chemical shift of the N1 atom in G537ins was shifted upfield relative to FMN in aqueous solution (Table 2). For pyridine-type nitrogen atoms, such shifts are reflective of hydrogen bonding at N1 of the oxidized flavin (36). Although pyrrole-type nitrogen atoms

such as N3 and N10 of flavin in the OX state are rather insensitive to changes in hydrogen bonding interactions, small downfield shifts do occur in response to such interactions (36). The 15N chemical shift for the N3 atom in the insertion variant shifted upfield by ∼2.6 ppm (to 157.9 ppm) relative to that of the wild type (Figure 4 and Table 2), reflecting a weaker hydrogen bonding interaction at the N3H group in G537ins. The N10 atom of G537ins resonates at 159.8 ppm, which was upfield by ∼2.8 ppm from those of FMN in an aqueous environment and wild-type FBDBM3, but downfield from TARF in CHCl3 (Table 2). Because the N10 atom cannot form hydrogen bonds, this upfield shift must be explained in some other manner. The shift could result from a decrease in the degree of sp2 hybridization if the atom was moved slightly out of the plane of the isoalloxazine ring (35) or from a change in the polarization of the isoalloxazine ring due to a weakening of the hydrogen bonding interactions at C2O and C4O (37). However, these groups interact with the backbone amide NH group of Gln579 and side chain hydroxyl group of Thr577 in the wild-type domain which are located on an adjacent loop (9). The upfield shift of the N10 resonance could also result from the disruption of the hydrogen bond at the N5 atom of the FMN (35, 37) as a result of the glycine insertion. Both the 15N-1H coupling constant and 15N chemical shift changes for the fully reduced state of wild-type FBDBM3 suggest a high degree of sp3 character of the N5H group due to the out-of-plane puckering of the central flavin ring, perhaps preventing the potential clash with the backbone amide group of Asn537 in the reduced enzyme (21). A similar investigation for the G537ins variant in the reduced state would be intriguing because of our hypothesis that this interaction was disrupted by the insertion. Unfortunately, we were unable to obtain satisfactory results for the reduced G537ins by 1D 15N NMR analyses due to very weak signals typically associated with 15N NMR analyses, the tendency for the sample to reoxidize and precipitate during extended data acquisition times. Alternatively, 1H-15N HSQC NMR studies on reduced G537ins were initiated because of its increased sensitivity, although this approach would be responsive to only the N3H and N5H moieties. Only the cross-peak for the N3H group of the reduced G537ins was observed (data not shown). The absence of a signal attributable to the N5H group was consistent with the 1D NMR results, again suggesting a rapid proton exchange with the solvent and/or signal broadening due to rapid changes in its environment. The temperature coefficient (∆δ/∆T) for the proton chemical shift has been used as an indicator of relative hydrogen bonding strength in that the chemical shift of the amide proton that is exposed to solvent is more sensitive to temperature than those involved in intermolecular hydrogen bonding (28, 38). This approach was applied here. The

13794

Biochemistry, Vol. 47, No. 52, 2008

temperature coefficient for N3H in reduced G537ins was found to be -4.4 ppb/K, a value similar to that for an amide proton involved in an intramolecular hydrogen bond (39), but that was 3-fold higher than that determined for the wild type under similar conditions (-1.65 ppb/K). These results indicate that the hydrogen bonding interaction at the N3H group was retained but significantly weakened by the glycine insertion, a conclusion consistent with the 1D NMR data. Unfortunately, the absence of an HSQC signal precluded a similar analysis for the N5H group. Molecular Modeling. To gain structural insights into how the inner FMN-binding loop of G537ins interacts with its flavin cofactor, a structural model of G537ins was generated. The X-ray crystal structure of wild-type FBDBM3 served as the initial template, and because of the general sequence homology of the two loops after the insertion, the FMNbinding domain of CPR was used as a guide (8, 9). As a starting point for the model, the backbone torsion angles of seven residues (534A-S-Y-G-N-G-H539) flanking the re face of the FMN isoalloxazine ring were initially adjusted to values similar to those of the corresponding residues in CPR with Swiss-PDB Viewer. The modeled loop was then subjected to multiple rounds of geometry optimization using the AMBER molecular mechanical force field and evaluated by structure validation software as described in Experimental Procedures. As a check, the initial structure of the loop was repeatedly adjusted and reoptimized to establish that the loop converged to the final structure reported here. Compared to the wild-type FBDBM3 structure, the final positions of the R-carbon atoms of the loop in the modeled structure of G537ins were more similar to those of the human CPR with a root-mean-square deviation (rmsd) of 0.83 Å (Figure 5A). The inserted glycine residue was situated at a position corresponding to Asn537 in the wild-type structure. Perhaps the most significant structural consequence of the insertion was the dislocation of the backbone amide nitrogen atom of Asn537 to a position that was at least 2.5 Å farther from the N5 atom of the flavin (Figure 5A). This displacement would certainly disrupt the hydrogen bond between Asn537 and the oxidized FMN that is observed in the wildtype structure (9). Furthermore, the amide NH group of the inserted glycine was positioned 3.7 Å from the N5 atom, a distance quite comparable to that of Gly141 in human CPR (3.8 Å) and too large for the formation of a new hydrogen bond at this position in the oxidized state, just as observed for human CPR and the flavodoxin (8). However, the carbonyl group of the inserted glycine was within a suitable distance to serve as a hydrogen bond acceptor for the N5H group in the reduced states of the FMN. The functional significance of this observation will be discussed below. The rest of the enlarged loop in G537ins appears to be stabilized by maintaining the two hydrogen bonds that are present between the backbone atoms of Tyr536 and His539 that help form a type I′ turn with Gly538 at the third position in the wild type (9). However, a different turn conformation was adopted that more closely resembles that of human CPR (Figure 5B). Two tandem proline residues, located at the end of the loop, remain in backbone conformations similar to that in the wild type. The side chain of Tyr536, which stacks on the re face of the flavin ring and hydrogen bonds with the phosphate group in wild-type FBDBM3 (9), was also in a

Chen and Swenson

FIGURE 5: (A) Peptide loop of the G537ins model (cyan) flanking the re face of the FMN cofactor (yellow) superimposed with the analogous regions of wild-type FBDBM3 [PDB entry 1bvy, F chain (orange)] and CPR [PDB entry 1b1c (blue)]. The structure of the loop region of G537ins closely resembles that of the human CPR. Note that Asn537 in the wild-type structure was pushed away from the N5 atom (highlighted in blue) in the modeled structure. Potential hydrogen bonds between N5 and the glycine residues in CPR (G141) and the modeled structure (G537*) are indicated by dashed lines. (B) Superposition of the residues 536Y-G-N-G538 in the modeled structure (cyan) and the corresponding residues in CPR (blue). The carbonyl group of the inserted glycine residue (G537*) in the modeled structure was oriented toward the FMN, and the potential hydrogen bond with the N5 atom of the cofactor is indicated by the dashed line. Both figures were generated using PyMOL (49).

similar position as in the wild-type structure, implying that these interactions can still be maintained in G537ins. Electron Transferring ActiVity of the DiflaVin Reductase Domain (BMR) Containing the G537 Insertion. The glycine insertion was initially introduced into just the FMN-binding domain to greatly facilitate the characterization of the effects of this alteration on the redox properties of the bound FMN, the principal goal of this study. However, this construction precluded the investigation of its effects on the electron transferring activities of this protein. The electron transferring characteristics of the diflavin reductase domain (BMR) of flavocytochrome P450BM-3 have been studied extensively

Glycine Insertion Modulates the Redox Properties of P450BM-3

Biochemistry, Vol. 47, No. 52, 2008 13795

Table 3: Steady-State Kinetic Measurements of the Wild-Type BMR and BMR-Glyins toward Electron Acceptors turnover number (min-1)a BMR-Glyins electron acceptor

wild type

without FMN

with FMN

CPRb

ferricyanide cytochrome c

7300 ( 370 4200 ( 330

2500 ( 120 11 ( 1

2000 ( 100c 1900 ( 130d

4300 2300

a Values were determined from at least three independent measurements under the assay conditions described in Experimental Procedures. b From ref 50. c Reactions were performed in the presence of a 2-fold molar excess of FMN over BMR-Glyins. d From the average maximal rate obtained under the conditions described in the legend of Figure 6.

and have been compared to those of the mammalian CPR. Thus, the glycine insertion was introduced at the level of BMR (i.e., BMR-Glyins) to assess its effects on the cytochrome reductase activity. Unlike for the wild-type BMR prepared in the same way, flavin analyses of the purified BMR-Glyins protein indicated that while nearly stoichiometric amounts of FAD were bound, nearly all the FMN cofactor was absent as purified. The loss of FMN in BMR-Glyins was anticipated on the basis of our observations of the loss of FMN during purification and the higher Kd determined for this variant at the FBDBM3 level (see above). Partial reconstitution of BMR-Glyins could be achieved by incubation of the purified protein with a 5-fold excess of FMN followed by dialysis. Flavin analysis revealed that BMR-Glyins could be partially reconstituted but retained only 0.17 mol of FMN per mole of enzyme (and FAD) under these conditions. These results are corroborated by the activity measurements below and are also consistent with the inability to fully reconstitute the FBDBM3 with FMN. The reasons for the incomplete reconstitution of the glycine insertion variants are not known at this time. Steady-state turnover measurements were initially performed for BMR-Glyins toward an external electron acceptor, cytochrome c, in the absence and the presence of free FMN. As purified, BMR-Glyins displayed 12 (data not shown). In the C. beijerinckii flavodoxin, the pKa of the neutral SQ has been estimated to be >13 (17). This substantial increase was attributed primarily to the strong hydrogen bonding interaction between the N5H group of the neutral FMNSQ after it was noted that the pKa was lowered by at least two pH units in the G57T variant which disrupts this interaction (17). The increase in the pKa of the FMNSQ in the G537ins variant can be rationalized in part by a disruption of the hydrogen bonding interaction between the amide NH group of Asn537 and the N5 atom of the FMN if preserved in the SQ state of the wild-type protein. The expansion of the loop would also result in a greater solvent exposure of the flavin ring. However, the rather dramatic increase in the pKa implied by the pH jump analyses implies that a new hydrogen bond is formed between the carbonyl group of the inserted glycine and the N5H group of the neutral SQ species as observed in the flavodoxin. Our molecular model does show that the carbonyl group of the inserted glycine residue is oriented toward the flavin and within hydrogen bonding distance of the N5H group (Figure 5). This interaction has been estimated to contribute up to 4.0 kcal/ mol to the stabilization of the relatively air stable neutral SQ species in the flavodoxin (28). However, the degree of stabilization in the G537ins variant appears to be lower as reflected by a narrower separation of the midpoint potentials between two one-electron couples and the lower levels of accumulation of the radical at equilibrium during reductive titrations. Just as for wild-type FBDBM3, the binding free energy for the neutral SQ state was only modestly higher than for the OX state (by 0.9 kcal/mol) (Table 1) compared to a differential stabilization of the neutral radical in the flavodoxin of 3.3 kcal/mol (28). These observations suggest that should a hydrogen bonding interaction be formed with the N5H group of the FMNSQ in G537ins, it is weaker than in the flavodoxin. This conclusion is supported by the 15N NMR evidence of a weaker hydrogen bond to the N3H group compared to the wild-type protein studies of the variant in both the oxidized and reduced states. In the flavodoxin, changes in interactions with the FMN N3H group can reflect those at the N5 atom (38). A somewhat surprising outcome was that the glycine insertion had the greatest effect on the stability of the FMNHQ. The midpoint potential for the SQ/HQ couple became substantially more negative than for the wild type and the OX/SQ couple in the variant, reflecting a loss of binding energy of ∼2.7 kcal/mol for the FMNHQ species relative to the wild type (Table 1). This observation is not easily explained, yet certain factors may contribute. At least a portion of this difference can be attributed to the reduction of the neutral SQ for G537ins rather than

Glycine Insertion Modulates the Redox Properties of P450BM-3 the anionic species in the wild-type protein. Reduction would be accompanied by differences in the charge distribution in the flavin isoalloxazine ring in each case. Computational studies suggest that the ESQ/HQ in the flavodoxin is not only dependent on the general electrostatic environment of the cofactor but also quite sensitive to the backbone configuration of the analogous loop in those proteins including the orientation of the backbone carbonyl group of the conserved glycine (48). The marked sensitivity of HQ state to even relatively subtle changes in the structure of the cofactor binding site has been noted previously in the flavodoxin (17). Therefore, altering the local backbone environment by expanding the size and structure of the loop in the FBDBM3 could certainly account for the lower stability of the HQ complex. The differences in redox properties of primarily the FMN cofactor in BM3 could form the basis for differing electron transferring mechanisms and/or regulatory phenomena (5, 22). As an initial effort to address some of these issues, steady-state electron transfer activities were evaluated for the glycine insertion introduced into the diflavin reductase domain (BMR) to assess its effect on electron transfer toward electron acceptors ferricyanide and cytochrome c and for comparison to mammalian CPR and other diflavin reductases. Our kinetic analyses were complicated by the observation that only a portion (up to 20%) of the recombinant BMR-Glyins protein was able to fully incorporate FMN to form the holoenzyme. Although the reason for this is not known, it is quite possible that not all of the protein is able to fold properly, particularly in the FMN domain region. FAD binding seems not to be impaired, however, as nearly stoichiometric levels of this cofactor were observed in the purified preparations. The functionality of the FAD domain was confirmed by the observed ferricyanide reductase activity in the insertion variant, albeit at a somewhat reduced level relative compared to the wild-type BMR. Although it is not understood, previous studies have suggested that the ability to transfer electrons from NADPH to FAD to ferricyanide might be affected by FMN deficiency in the FMN-depleted P450BM-3 mutants, which is the case here (30). The evaluation of the cytochrome c reductase activity was made more challenging by of the incomplete reconstitution of the BMR-Glyins protein with the FMN cofactor, which is the primary donor to the cytochrome. However, the turnover number for the holoprotein that could be fully constituted was established from the maximal activity achieved under limiting levels of added FMN (Figure 6). The turnover number obtained in this manner, 1900 ( 130 min-1, indicates that the insertion variant retained approximately half of the wild-type cytochrome c reductase activity, which was more comparable to that of CPR obtained under similar assay conditions (2300 min-1) (50). It is generally accepted that only the anionic FMNSQ is capable of delivering an electron to the cytochrome acceptor in BM3 and BMR (12). The more thermodynamically favored FMNHQ is believed not to be kinetically competent to do so, forming at a rate slower than the rate of reduction of cytochrome acceptors (13). The FMNSQ formed in mammalian CPRs, in this case as the thermodynamically stable neutral form, serves as the electron donor to the cytochrome (5). Conversion of BMR into a

Biochemistry, Vol. 47, No. 52, 2008 13797

reductase displaying more CPR-like properties for the FMN cofactor was not, therefore, expected to preclude electron transfer since both types of reductases are capable of doing so. Thus, the alterations of the protonation state and the relative stability of the FMNSQ or the midpoint potentials of the FMN in BMR-Glyins did not appreciably affect the overall electron transfer activity toward this acceptor. However, other steps in the catalytic pathway are thought to be rate-limiting, including NADPH binding and hydride ion transfer from NADPH to FAD, etc. (5), which likely mask changes in the inter-flavin and/or cytochrome electron transfer steps. Although not fully understood for either reductase, the mechanisms and kinetics of electron transfer to their physiological redox partners will likely be different from those determined for cytochrome c reduction. It seems plausible that the differing properties of the FMN cofactor and its binding site in BM3 have evolved for specific purposes in this enzyme and the insertion variant will need to be generated and studied at the monooxygenase level. Such studies are being initiated. In conclusion, the results of this study support the hypothesis that the shorter loop that flanks the re face of the FMN in flavocytochrome P450BM-3 promotes the formation of a strong hydrogen bond to the N5 atom of the FMN in the oxidized state and that this interaction was maintained upon reduction to the anionic SQ. The expansion of this loop by the insertion of a glycine residue at a position where such a residue is highly conserved disrupted this interaction and promoted the formation and accumulation of the neutral form of SQ as observed in CPR, the flavodoxins, and other flavoproteins. These results also support our hypothesis on the importance of the structural role of the FMN-binding loop underlying the unique redox properties of P450BM-3 compared to those of CPR. Our functional studies suggest that the insertion within the reductase (BMR) does not significantly affect its general electron transfer activity. However, additional experiments in the full-length P450BM-3 protein are needed to improve our understanding of its effect on the electron transfer during fatty acid hydroxylation. Finally, the results of this study further expand our understanding of the specific structural features that are responsible for establishing the redox properties of flavoproteins. ACKNOWLEDGMENT We thank Dr. Chunhua Yuan of the Campus Chemical Instrument Center for assisting in NMR data collection. REFERENCES 1. Munro, A. W., Leys, D. G., McLean, K. J., Marshall, K. R., Ost, T. W., and Daff, S. (2002) P450 BM3: The very model of a modern flavocytochrome. Trends Biochem. Sci. 27, 250–257. 2. Yun, C. H., Kim, K. H., Kim, D. H., Jung, H. C., and Pan, J. G. (2007) The bacterial P450 BM3: A prototype for a biocatalyst with human P450 activities. Trends Biotechnol. 25, 289–298. 3. Narhi, L. O., and Fulco, A. J. (1986) Characterization of a catalytically self-sufficient 119,000-dalton cytochrome P-450 monooxygenase induced by barbiturates in Bacillus megaterium. J. Biol. Chem. 261, 7160–7169. 4. Ruettinger, R. T., Wen, L. P., and Fulco, A. J. (1989) Coding nucleotide, 5′ regulatory, and deduced amino acid sequences of

13798

5. 6. 7.

8.

9. 10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21. 22.

Biochemistry, Vol. 47, No. 52, 2008

P-450BM-3, a single peptide cytochrome P-450:NADPH-P-450 reductase from Bacillus megaterium. J. Biol. Chem. 264, 10987– 10995. Murataliev, M. B., Feyereisen, R., and Walker, F. A. (2004) Electron transfer by diflavin reductases. Biochim. Biophys. Acta 1698, 1–26. Porter, T. D. (1991) An unusual yet strongly conserved flavoprotein reductase in bacteria and mammals. Trends Biochem. Sci. 16, 154– 158. Porter, T. D., and Kasper, C. B. (1986) NADPH-cytochrome P-450 oxidoreductase: Flavin mononucleotide and flavin adenine dinucleotide domains evolved from different flavoproteins. Biochemistry 25, 1682–1687. Zhao, Q., Modi, S., Smith, G., Paine, M., McDonagh, P. D., Wolf, C. R., Tew, D., Lian, L. Y., Roberts, G. C., and Driessen, H. P. (1999) Crystal structure of the FMN-binding domain of human cytochrome P450 reductase at 1.93 Å resolution. Protein Sci. 8, 298–306. Sevrioukova, I. F., Li, H., Zhang, H., Peterson, J. A., and Poulos, T. L. (1999) Structure of a cytochrome P450-redox partner electrontransfer complex. Proc. Natl. Acad. Sci. U.S.A. 96, 1863–1868. Wang, M., Roberts, D. L., Paschke, R., Shea, T. M., Masters, B. S., and Kim, J. J. (1997) Three-dimensional structure of NADPHcytochrome P450 reductase: Prototype for FMN- and FADcontaining enzymes. Proc. Natl. Acad. Sci. U.S.A. 94, 8411–8416. Munro, A. W., Noble, M. A., Robledo, L., Daff, S. N., and Chapman, S. K. (2001) Determination of the redox properties of human NADPH-cytochrome P450 reductase. Biochemistry 40, 1956–1963. Sevrioukova, I., Shaffer, C., Ballou, D. P., and Peterson, J. A. (1996) Equilibrium and transient state spectrophotometric studies of the mechanism of reduction of the flavoprotein domain of P450BM-3. Biochemistry 35, 7058–7068. Murataliev, M. B., Klein, M., Fulco, A., and Feyereisen, R. (1997) Functional interactions in cytochrome P450BM3: Flavin semiquinone intermediates, role of NADP(H), and mechanism of electron transfer by the flavoprotein domain. Biochemistry 36, 8401–8412. Watt, W., Tulinsky, A., Swenson, R. P., and Watenpaugh, K. D. (1991) Comparison of the crystal structures of a flavodoxin in its three oxidation states at cryogenic temperatures. J. Mol. Biol. 218, 195–208. Zhou, Z., and Swenson, R. P. (1996) The cumulative electrostatic effect of aromatic stacking interactions and the negative electrostatic environment of the flavin mononucleotide binding site is a major determinant of the reduction potential for the flavodoxin from DesulfoVibrio Vulgaris [Hildenborough]. Biochemistry 35, 15980– 15988. O’Farrell, P. A., Walsh, M. A., McCarthy, A. A., Higgins, T. M., Voordouw, G., and Mayhew, S. G. (1998) Modulation of the redox potentials of FMN in DesulfoVibrio Vulgaris flavodoxin: Thermodynamic properties and crystal structures of glycine-61 mutants. Biochemistry 37, 8405–8416. Ludwig, M. L., Pattridge, K. A., Metzger, A. L., Dixon, M. M., Eren, M., Feng, Y., and Swenson, R. P. (1997) Control of oxidation-reduction potentials in flavodoxin from Clostridium beijerinckii: The role of conformation changes. Biochemistry 36, 1259–1280. Smith, W. W., Burnett, R. M., Darling, G. D., and Ludwig, M. L. (1977) Structure of the semiquinone form of flavodoxin from Clostridum MP. Extension of 1.8 Å resolution and some comparisons with the oxidized state. J. Mol. Biol. 117, 195–225. Barsukov, I., Modi, S., Lian, L. Y., Sze, K. H., Paine, M. J., Wolf, C. R., and Roberts, G. C. (1997) 1H, 15N and 13C NMR resonance assignment, secondary structure and global fold of the FMNbinding domain of human cytochrome P450 reductase. J. Biomol. NMR 10, 63–75. Garcin, E. D., Bruns, C. M., Lloyd, S. J., Hosfield, D. J., Tiso, M., Gachhui, R., Stuehr, D. J., Tainer, J. A., and Getzoff, E. D. (2004) Structural basis for isozyme-specific regulation of electron transfer in nitric-oxide synthase. J. Biol. Chem. 279, 37918–37927. Kasim, M. (2002) Ph.D. Thesis, The Ohio State University, Columbus, OH. Daff, S. N., Chapman, S. K., Turner, K. L., Holt, R. A., Govindaraj, S., Poulos, T. L., and Munro, A. W. (1997) Redox control of the catalytic cycle of flavocytochrome P-450 BM3. Biochemistry 36, 13816–13823.

Chen and Swenson 23. Rock, D., and Jones, J. P. (2001) Inexpensive purification of P450 reductase and other proteins using 2′,5′-adenosine diphosphate agarose affinity columns. Protein Expression Purif. 22, 82–83. 24. Marohnic, C. C., Panda, S. P., Martasek, P., and Masters, B. S. (2006) Diminished FAD binding in the Y459H and V492E AntleyBixler syndrome mutants of human cytochrome P450 reductase. J. Biol. Chem. 281, 35975–35982. 25. Clark, W. M. (1960) Oxidation-reduction potentials of organic systems, Williams & Wilkins, Baltimore. 26. Haines, D. C., Sevrioukova, I. F., and Peterson, J. A. (2000) The FMN-binding domain of cytochrome P450BM-3: Resolution, reconstitution, and flavin analogue substitution. Biochemistry 39, 9419–9429. 27. Dubourdieu, M., le Gall, J., and Favaudon, V. (1975) Physicochemical properties of flavodoxin from DesulfoVibrio Vulgaris. Biochim. Biophys. Acta 376, 519–532. 28. Chang, F. C., and Swenson, R. P. (1999) The midpoint potentials for the oxidized-semiquinone couple for Gly57 mutants of the Clostridium beijerinckii flavodoxin correlate with changes in the hydrogen-bonding interaction with the proton on N(5) of the reduced flavin mononucleotide cofactor as measured by NMR chemical shift temperature dependencies. Biochemistry 38, 7168– 7176. 29. Mori, S., Abeygunawardana, C., Johnson, M. O., and van Zijl, P. C. (1995) Improved sensitivity of HSQC spectra of exchanging protons at short interscan delays using a new fast HSQC (FHSQC) detection scheme that avoids water saturation. J. Magn. Reson., Ser. B 108, 94–98. 30. Klein, M. L., and Fulco, A. J. (1993) Critical residues involved in FMN binding and catalytic activity in cytochrome P450BM-3. J. Biol. Chem. 268, 7553–7561. 31. Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) PROCHECK: A program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26, 283–291. 32. Vriend, G. (1990) WHAT IF: A molecular modeling and drug design program. J. Mol. Graphics 8, 52–56. 33. Nishimoto, K., Watanabe, Y., and Yagi, K. (1978) Hydrogen bonding of flavoprotein. I. Effect of hydrogen bonding on electronic spectra of flavoprotein. Biochim. Biophys. Acta 526, 34–41. 34. Sevrioukova, I., Truan, G., and Peterson, J. A. (1996) The flavoprotein domain of P450BM-3: Expression, purification, and properties of the flavin adenine dinucleotide- and flavin mononucleotide-binding subdomains. Biochemistry 35, 7528–7535. 35. Vervoort, J., Muller, F., Mayhew, S. G., van den Berg, W. A., Moonen, C. T., and Bacher, A. (1986) A comparative carbon-13, nitrogen-15, and phosphorus-31 nuclear magnetic resonance study on the flavodoxins from Clostridium MP, Megasphaera elsdenii, and Azotobacter Vinelandii. Biochemistry 25, 6789–6799. 36. Witanowski, M., Stefaniak, L., and Webb, G. A. (1981) Nitrogen NMR spectroscopy. Annu. Rep. NMR Spectrosc. 11B, 1–493. 37. Moonen, C. T. W., Vervoort, J., and Muller, F. (1984) Reinvestigation of the structure of oxidized and reduced flavin: Carbon-13 and nitrogen-15 nuclear magnetic-resonance study. Biochemistry 23, 4859–4867. 38. Bradley, L. H., and Swenson, R. P. (2001) Role of hydrogen bonding interactions to N(3)H of the flavin mononucleotide cofactor in the modulation of the redox potentials of the Clostridium beijerinckii flavodoxin. Biochemistry 40, 8686–8695. 39. Cierpicki, T., and Otlewski, J. (2001) Amide proton temperature coefficients as hydrogen bond indicators in proteins. J. Biomol. NMR 21, 249–261. 40. Paine, M. J. I., Garner, A. P., Powell, D., Sibbald, J., Sales, M., Pratt, N., Smith, T., Tew, D. G., and Wolf, C. R. (2000) Cloning and characterization of a novel human dual flavin reductase. J. Biol. Chem. 275, 1471–1478. 41. Leclerc, D., Wilson, A., Dumas, R., Gafuik, C., Song, D., Watkins, D., Heng, H. H. Q., Rommens, J. M., Scherer, S. W., Rosenblatt, D. S., and Gravel, R. A. (1998) Cloning and mapping of a cDNA for methionine synthase reductase, a flavoprotein defective in patients with homocystinuria. Proc. Natl. Acad. Sci. U.S.A. 95, 3059–3064. 42. Sibille, N., Blackledge, M., Brutscher, B., Coves, J., and Bersch, B. (2005) Solution structure of the sulfite reductase flavodoxinlike domain from Escherichia coli. Biochemistry 44, 9086–9095. 43. Sharkey, C., Mayhew, S. G., Higgins, T. M., and Walsh, M. A. (1997) in FlaVins and FlaVoproteins 1996 (Stevenson, K. J.,

Glycine Insertion Modulates the Redox Properties of P450BM-3 Massey, V., and Williams, C. H., Jr., Eds.) pp 445-448, University of Calgary Press, Calgary, AB. 44. Romero, A., Caldeira, J., Legall, J., Moura, I., Moura, J. J., and Romao, M. J. (1996) Crystal structure of flavodoxin from DesulfoVibrio desulfuricans ATCC 27774 in two oxidation states. Eur. J. Biochem. 239, 190–196. 45. Luschinsky, C. L., Dunham, W. R., Osborne, C., Pattriege, K. A., and Ludwig, M. L. (1991) in FlaVins and FlaVoproteins 1990 (Curti, B., Ronchi, S., and Zanetti, G., Eds.) pp 409-413, W. de Gruyter, Berlin. 46. Kasim, M., and Swenson, R. P. (2000) Conformational energetics of a reverse turn in the Clostridium beijerinckii flavodoxin is directly coupled to the modulation of its oxidation-reduction potentials. Biochemistry 39, 15322–15332.

Biochemistry, Vol. 47, No. 52, 2008 13799

47. Draper, R. D., and Ingraham, L. L. (1968) A potentiometric study of the flavin semiquinone equilibrium. Arch. Biochem. Biophys. 125, 802–808. 48. Ishikita, H. (2008) Redox potential difference between DesulfoVibrio Vulgaris and Clostridium beijerinckii flavodoxins. Biochemistry 47, 4394–4402. 49. DeLano, W. L. (2002) The PyMol Molecular Graphics System, DeLano Scientific, San Carlos, CA. 50. Kurzban, G. P., and Strobel, H. W. (1986) Preparation and characterization of FAD-dependent NADPH-cytochrome P-450 reductase. J. Biol. Chem. 261, 7824–7830.

BI800954H